Every conversation in this industry right now seems to start the same way: “What AI tools should I be using?” I believe that’s the wrong question.
After spending the past year building and testing AI applications inside our business using real transactions with buyers and sellers, I’ve learned that the tools matter far less than the workflow you point them at.
For brokerages looking to move from experimenting with AI to actually implementing it, the best place to start is with the parts of your business where you’re losing time, leads or opportunities. Here are five ways to do that.
How to implement AI in your business
1. Start by identifying the friction in your business
Before buying another AI product, map out your brokerage’s core workflows: lead intake, nurture, property search, listing marketing, offer preparation and transaction management.
Then ask three questions: Where are agents spending too much time? Where are consumers dropping off? Where is repetitive work slowing down the transaction?
Pick one or two of those problems to solve first. Don’t start with a list of AI tools and try to find places to use them. Start with a measurable business problem and determine whether AI can improve it.
Action to take: Choose one workflow this month and document every step from beginning to end. Identify the two or three places where your team loses the most time or where clients experience the most friction. Those are your first AI use cases.
2. Turn lead generation into lead intelligence
The industry has spent two decades obsessing over generating more leads. But for many brokerages, the bigger opportunity is understanding the leads they already have.
Traditional portals force consumers into rigid filters: bedrooms, bathrooms, ZIP code and price. But that’s not how people actually think about where they want to live.
Conversational AI can capture much more nuanced intent. A prospect might say, “I want a contemporary house, 20 minutes from Bethesda, room for a pool, great entertaining space, under $2 million.” Instead of reducing that person to five search filters, AI can begin building a much richer picture of what they actually want.
That’s one of the areas I’ve focused on with an AI platform I’ve been developing around intent-based search. But the larger opportunity applies to any brokerage: Use AI to understand your prospects before simply adding more names to your database.
Action to take: Look at your current lead intake process. Beyond price, location and property type, start capturing lifestyle preferences, motivations, timing and deal-breakers. Feed that information into your CRM or AI workflow so agents receive a meaningful client profile, not just another lead notification.
3. Give every agent an AI communication assistant
One of the most immediate applications of AI is also one of the simplest: Use it to eliminate repetitive communication and marketing work.
AI can help agents draft personalized follow-up emails, listing descriptions, social posts, neighborhood content, buyer summaries, showing follow-ups and seller presentations. It can also make lead nurture more conversational than the traditional automated drip campaign.
We’ve seen consumers engage differently when the experience feels like a conversation rather than a generic marketing sequence. They can share what they’re looking for, ask questions and provide more context before an agent ever schedules the first showing.
The goal isn’t to remove agents from communication. It’s to let technology handle the repetitive work around the relationship so agents can spend more time building trust, advising clients and closing deals.
Action to take: Build a small library of approved AI prompts and templates around the five communications your agents write most often. Train agents to personalize the output rather than simply copying and pasting it, and measure how much time it saves each week.
4. Bring AI into the transaction itself
Don’t stop using AI once a lead becomes a client. Some of the biggest efficiency gains can happen after the deal starts.
I’m currently developing a second application that can draft an offer in seconds and deliver it with signatures ready to go. The objective is simple: Spend less time assembling paperwork and more time advising the client.
Brokerages can apply that same thinking throughout the transaction. AI can summarize inspection reports, compare properties, analyze comps, identify questions within lengthy documents, create transaction checklists, and translate complicated paperwork into plain language or a client’s preferred language.
Action to take: Identify the three most document-heavy tasks your agents or transaction teams perform repeatedly. Test AI on those tasks with appropriate human review and compliance safeguards. Track how long the process takes today versus with AI assistance.
5. Build an AI operating system, not a tool graveyard
This is where I see many brokerages get AI adoption wrong. They treat it like a shopping list, subscribing to the newest product every time one launches and eventually ending up with 20 disconnected tools that don’t talk to each other.
Instead, establish a simple framework before adopting anything new.
- What problem does this solve?
- Who will use it?
- What data does it need?
- Does it integrate with the systems we already use?
- What does success look like?
- And how will we measure ROI?
If you can’t answer those questions, you probably don’t need the tool.
Action to take: Audit every AI product your brokerage is currently paying for. Assign each one a specific workflow, owner and measurable outcome. If a product doesn’t have all three, reconsider whether it belongs in your technology stack.
Start with one workflow, then scale
Real estate has spent 20 years digitizing much of the same underlying process. We took the MLS book, put it online, added filters, better photos and mobile apps, but much of the consumer and agent experience still works fundamentally the same way.
AI gives brokerages an opportunity to rethink those workflows rather than simply digitize them again. But you don’t need to transform the entire company overnight.
Start with one problem. Establish a baseline. Test an AI-enabled workflow. Measure the results. Keep what works, fix what doesn’t and then move to the next problem.
If there’s one question I’d want every broker and team leader to ask, it’s not, “How can I use AI?”
Ask instead: “What part of my business should no longer work the way it works today?”
That’s where I would start.
This October, Inman turns its focus to AI and its rapid rise in real estate. During Artificial Intelligence Month, we’re digging into the startups shaking things up and the established players folding AI into their offerings.
Nurit Coombe is a team leader at The Agency and co-founder of HouseMe.ai, an AI-powered home search platform. Get connected on Instagram and LinkedIn.
